The primary purpose of the research in this project is to specify and estimate models for dietary intakes of the subjects in the Women's Health Trial: Feasibility Study in Minority Populations (WHTFSMP) taking into account the behavioral factors emphasized in the Social Learning Theory and the Health Belief Model. The comprehensive analysis will incorporate the socioeconomic factors such as education and income, and the physiological aspects such as anthropometric indicators and the current energy intakes that reflect the subjects' energy requirements. In addition, the theories of "habit persistence" in diets from the economics literature will be incorporated in the modeling framework. The subjects in the Intervention group of the WHTFSMP received counseling for adopting diets that were low in fat and high in fruits and vegetables. The research will analyze the proximate determinants of the intakes of carbohydrate, saturated, monounsaturated, and polyunsaturated fats, fiber, calcium, a-carotene and ascorbic acid using the data at baseline, six and 12 months from the Control and Intervention groups of the WHTFSMP. The second objective of the research is to estimate dynamic random effects models for the subjects' risk factors for coronary disease and cancer such as plasma LDL and HDL cholesterol, triglycerides, and glucose levels. Behavioral factors, anthropometric indicators, energy derived from saturated and polyunsaturated fats, and dietary cholesterol will be explanatory variables in models for plasma LDL cholesterol. The dynamic model is suitable for plasma cholesterol since 50% of cholesterol is endogenously produced. The comprehensive analysis of the data from the WHTFSMP will be useful for refining educational programs that seek to promote healthful eating. In particular, the incorporation of the various behavioral factors in the analyses of dietary intakes will identify strategies of targeting women who are at greater risk of chronic disease but are less likely to change their behavior and would benefit from enrollment in nutrition education programs. The results from the empirical models for plasma LDL, HDL, triglycerides and glucose will be useful for identifying strategies for lowering the subjects' risks for cardiovascular disease and cancers. Overall, the results will facilitate the design of policies for improving health of especially minority women in the United States and will be useful for analyses of the Women's Health Initiative data.